Panoptic-DeepLab (SWideRNet [1, 1, 4.5], Mapillary, multi-scale)

Reported on 3 benchmarks across 3 tasks · 1 paper · 3 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Medical1 result

Semantic SegmentationonCityscapes test
PQ· uses extra data· 2020-11-23
67.8
best: 68 (OneFormer (ConvNeXt-L, single-scale, Mapillary Vistas-Pretrained))
SOTA
Scaling Wide Residual Networks for Panoptic Segmentation arXiv:2011.11675

Audio1 result

10-shot image generationonCityscapes test
PQ· uses extra data· 2020-11-23
67.8
best: 68 (OneFormer (ConvNeXt-L, single-scale, Mapillary Vistas-Pretrained))
SOTA
Scaling Wide Residual Networks for Panoptic Segmentation arXiv:2011.11675

Computer Vision1 result

Panoptic SegmentationonCityscapes test
PQ· uses extra data· 2020-11-23
67.8
best: 68 (OneFormer (ConvNeXt-L, single-scale, Mapillary Vistas-Pretrained))
SOTA
Scaling Wide Residual Networks for Panoptic Segmentation arXiv:2011.11675